Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://geo.web.ru/~ariskin/docs/papers/p97_046.pdf Äàòà èçìåíåíèÿ: Fri Jan 31 12:10:58 2003 Äàòà èíäåêñèðîâàíèÿ: Sat Dec 22 04:31:26 2007 Êîäèðîâêà: Windows-1251

Petrology, Vol. 5, No. 1, 1997, pp. 28-36. Translated from Petrologiya, Vol. 5, No. 1, 1997, pp. 32-41. Original Russian Text Copyright ¿ 1997 by Ariskin, Meshalkin, Almeev, Barmina, Nikolaev. English Translation Copyright ¿ 1997 by åÄàä çÛÍ / Interperiodica Publishing (Russia).

INFOREX Information Retrieval System: Analysis and Processing of Experimental Data on Phase Equilibria in Igneous Rocks
A. A. Ariskin*, S. S. Meshalkin*, R. R. Almeev**, G. S. Barmina*, and G. S. Nikolaev*
* Vernadsky Institute of Geochemistry and Analytical Chemistry, Russian Academy of Sciences, ul. Kosygina 19, Moscow, 117975 Russia ** Department of Geology, Moscow State University, Vorob'evy Gory, Moscow, 119899 Russia
Received April 1, 1995

Abstract--The most recent version of the INFOREX database (INFOREX 4.0, 1996), which includes experimental data on crystal-melt equilibria and a package of programs for their analysis and processing, is presented. The database contains the results of 186 studies (about 7000 experiments and more than 9600 phase compositions) and is currently the most complete information system in the field of experimental petrology of igneous rocks. The main options of INFOREX can be divided into the information, retrieval, and applied (Petrological Calculations) categories. The computational functions include: (1) calculation of regression coefficients describing mineral-melt equilibria for olivine, plagioclase, three varieties of pyroxenes and spinellids: a total of 28 geothermometers; (2) calculation of the Fe3+/Fe2+ ratio in experimental glasses for nine models; (3) calculation of the H2O solubility in melts for water-saturated experiments for three models; and (4) auxiliary calculations, including the projection of glass compositions onto ternary diagrams. The paper presents examples of using the INFOREX system to analyze the high-pressure cotectic melts Ol-Opx-Cpx and equilibrium isotherms between melts and amphibole under water-saturated conditions.

INTRODUCTION Numerous experimental studies of phase equilibria in synthetic and natural silicate systems have yielded enormous amounts of data on the conditions of quench experiments and the phase compositions of their products. According to our estimates, the overall number of subliquidus high-temperature experiments alone is at least 10 000; these are approximately equaled in number by microprobe analyses of the phases (crystals and quench glasses) synthesized in these experiments. In order to make efficient use of this vast experimental information in petrological and geochemical investigations, it is necessary to analyze hundreds of publications and assess the reliability of the data presented therein. This problem is particularly urgent in computer modeling of magmatic equilibria based on calibrations that simultaneously include data from hundreds and thousands of experiments (Nielsen, 1990; Weaver and Langmuir, 1990; Ariskin et al., 1993; Ghiorso and Sack, 1995). Furthermore, petrological reconstructions rely heavily on phase diagrams, geothermobarometers, and correlations, which must incorporate experimental data so as to compare empirical relationships and extrapolations proposed by various workers and to evaluate their accuracy. To accomplish these tasks, it is essential to have rapid access to the available information and to be able to filter the data by major-element criteria (magma composition) and experimental conditions, including duration, temperature, pressure, f O2
28

and volatile conditions. Apparently, a resolution of these issues could be found in a global computer system that would accumulate the experimental information available on phase equilibria of igneous rocks and provide an opportunity to assess the data for reliability. A few years ago, we started working on such a system (Ariskin et al., 1992) and recently presented the INFOREX 3.0 database developed for IBM-compatible personal computers running MS-DOS (Meshalkin and Ariskin, 1996; Ariskin et al., 1996). In addition to experimental information, this package includes a userfriendly multifunctional Data Base Management System (DBMS) (Meshalkin et al., 1996), which proved to be very efficient in practice (Ariskin and Nikolaev, 1995; Almeev and Ariskin, 1996; Ariskin and Nikolaev, 1996). This prompted us to integrate a number of new functions into the user interface for solving some applied petrological problems. This paper will give the most complete description of the experimental data set included in the INFOREX version 4.0 (1996) and a few examples of its use. BRIEF DESCRIPTION OF THE INFOREX SYSTEM Here, the term system is used to refer to both the database and the built-in module that processes the user-requested information retrieved from the base. This may be regarded as the first and essential step toward creating a petrological expert system that can


INFOREX INFORMATION RETRIEVAL SYSTEM

29

both analyze experimental data and use the results to simulate phase equilibria and test various computergenerated models. File Structure and Hierarchy The INFOREX database is based on text files with tabulated compositions of experimentally synthesized phases. The top-level files hold references and run conditions. The master file, containing experimental conditions (CONDIT.EXP) (Meshalkin et al., 1996), is divided into fixed fields that include: reference number, experiment number, rock name, address of the starting material composition, flag for the presence or absence of volatile components, pressure, temperature, oxygen fugacity, experiment duration, and container type. The last seven fields serve to identify the phase assemblages (glass, olivine, garnet, etc.), with flags indicating the availability of microprobe analyses. With certain exceptions (for sulfides, carbonates, and metals), the composition files are formatted uniformly and contain the contents of 12 major oxides in the order: SiO2 , TiO2 , Al2O3 , FeO (total), MnO, MgO, CaO, Na2O, K2O, P2O5 , Cr2O3 , and H2O. A separate file stores additional information on hydrous experiments: compositions of the volatile components and, if present, of the fluid phases (H2O, CO2 , and H2) in weight and mole percent. Statistics of Experimental Data As of April 1996, the database contains information reported in 186 works, such as papers, books, and dissertations, between 1962 and 1995, as well as some unpublished information that was available to us (Fig. 1a). Of the 6900 experiments incorporated into the database, more than 80% relate to natural rocks; 4580 experiments were performed at dry conditions, and the rest were conducted in the presence of H2O and/or CO2. The high-pressure experiments account for no more than 40%. Figures 1c and 1d show the distribution of the experiments as a function of pressure and temperature. The total number of microprobe-determined phase compositions is 9618, and the most representative data are available for quench glasses (3634), olivine (1442), plagioclase (1026), and high-calcium pyroxene (914) (Fig. 1b). A special data set of the INFOREX system includes 298 glass compositions with known ferric-ferrous ratios (Nikolaev et al., 1996). For 94 hydrous experiments, direct water determinations are given, in addition to the major-element concentrations in quench glasses (Almeev and Ariskin, 1996). Main DBMS Options The concept of the INFOREX DBMS consists in the idea that we create not so much a "store" of some potentially useful information as a working system
PETROLOGY Vol. 5 No. 1 1997

designed to solve both traditional and current petrological problems. In the future, this system may become a part of a more general package that, in addition to the database and retrieval software, would include various computer models of phase equilibria applicable to closed and open geochemical systems. Thus, the main DBMS options can be classified into the information (1, 2, 3, and 6), searching (4 and 5), and applied (7) categories: (1) System Configuration; (2) General System Information; (3) View / Edit / Input Data Files; (4) Selecting Experimental Conditions; (5) Setting Phase Assemblages; (6) Exchange INFOREX Data; and (7) Petrological Calculations. The first six options were detailed in our previous papers (Meshalkin et al., 1996; Meshalkin and Ariskin, 1996; Ariskin et al., 1996), so here we shall only describe several new features that have been added to the DBMS during the last few months. These include particularly new capabilities of the search procedure, which operates in two stages that involve the selection of run conditions (4) and the definition of the set of liquidus phases (5). Option (4) has now been supplemented so that it is possible to choose water-saturated and/or water-undersaturated experiments and to save records containing water determinations in melts to a separate file; an analogous search is possible for experiments with known ferric-ferrous ratios in quench glasses. Additionally, the glass compositions are now constrained not only by the oxide contents (SiO2 , Al2O3 , CaO, Na2O + K2O, TiO2 , FeO, and MgO) and characteristic ratios (FeO/MgO and Mg/(Mg + Fe)), but also by dividing the data into three magmatic series (tholeiitic, subalkaline, and alkaline) based on the (Na2O + K2O)-SiO2 diagram shown in Fig. 2 (Klassifikatsiya i nomenklatura ..., 1981). When setting the phase assemblages [option (5)], it is now possible to specify more exactly the set of liquidus phases, selecting only those runs that have glass compositions in equilibrium with a desired mineral (e.g., plagioclase: Pl-Liq, Ol-Pl-Liq, Ol-Pl-Cpx-Liq, etc.), with a mineral assemblage (e.g., Ol-Opx-Cpx-Liq), or experiments involving the compositions of two desired phases (olivine-melt, orthopyroxene-garnet, etc.). This additional capability is geared to the needs of petrologists developing mineral-melt geothermometers, two-mineral geobarometers, or attempting to classify the compositions of liquids produced by partial melting of peridotite. Despite the large amount of information analyzed, the search procedures of the INFOREX DBMS run on a PC of 486 class or higher and take only a few seconds. The search results are saved as conveniently usable files formatted in tabular form, which can be easily integrated into standard graphic software.


30 16 N = 186 (a)

ARISKIN et al. 4000 (b) Number of compositions 3000

14 Number of studies 12 10 8 6 4 2

2000

1000 0

70 % experiments 60 % experiments 50 40 30 20 10 0 0 562 232 37 50 31 100 1670 4143 exper.

60 2747 exper. 40

20 (c) 16 % experiments 692 12 8 4 0 600 n = 6900 (d)

20 0

138 309

356 4 6 kbar 7 250

116 8 10

1 atm 2 24 23

11

150 200 Pressure, kbar

300

Fig. 1. Statistics of experimental data available in the INFOREX 4.0 (1996) database. Distribution of the number of studies in terms of year of publication (a), phase compositions (b), and experimental pressure (c) and temperature (d). Phase symbols are as used in the database (Meshalkin et al., 1996).

PETROLOGICAL CALCULATIONS In the research conducted by our group, the principal function of the INFOREX system is to prepare data for developing mineral-melt geothermometers with the subsequent integration of these relationships into the COMAGMAT package of petrological programs, designed to simulate phase equilibria in magmatic melts (Ariskin et al., 1993). At the same time, the solution of this specific task frequently involves a number of practical petrological tasks concerning the division of iron into the ferric and ferrous oxidation states, estimation of oxygen fugacity and water content in the melt, various major-element recalculations, including projection of glass compositions onto pseudoternary diagrams, construction of concentration plots, and so on. The need for additional programming aimed at automating all these procedures led to the creation of a special routine of the INFOREX database called Petrological Calculations. Its structure and main functions are shown in Fig. 3. The Petrological Calculations include: (1) Calculation of Regression Coefficients describing mineral-

melt equilibria for olivine, plagioclase, three varieties of pyroxenes and spinellids: a total of 28 geothermometers; (2) Calculation of the Fe3+/Fe2+ Ratio in experimental glasses according to nine models (Sack et al., 1980; Kilinc et al., 1983; Kress and Carmichael, 1988; Borisov and Shapkin, 1989; Kress and Carmichael, 1991; Mysen, 1991; and three models in Nikolaev et al., 1996); (3) Calculation of H2O Solubility in melts for water-saturated experiments according to three models (Burnham, 1994; Dixon et al., 1995; Almeev and Ariskin, 1996); and (4) Auxiliary Calculations, including the normalization of glass compositions to 100% (for dry and hydrous systems) and their projection onto ternary diagrams, such as Ol-Pl-Qtz and Qtz-An-Ab. These functions require the same procedure for all types of calculations: at first, the user browses through the list of search-generated initial data files, selects one such file, and sets the type of calculation (1-4); if necessary, the model parameters are specified. After that, control is transferred to the program, which creates an output file. The user can view the computational results
PETROLOGY Vol. 5 No. 1 1997

LIQ OLIV PLAG AUG OPX SPIN PIG MAGN HORN GARN ILM LEUC ALFS MEL MICA NEPH AMPH 900 1200 1500 1800 Temperature, ϑ 2100 2400

0 1960 1965 1970 1975 1980 1985 1990 1995 Years


INFOREX INFORMATION RETRIEVAL SYSTEM CaO, wt % 14 Ol + Pl + Cpx + Spl + Liq N = 35 12 10 21 8 17 N = 343 13 9 5 1 30 40 50 60 70 80 SiO2, wt % N = 769 5 6 7 8 9 MgO, wt %

31

Na2O + K2O, wt % 25

N = 2522

Fig. 2. Division of glass compositions into conventional tholeiitic, subalkaline, and alkaline series (Klassifikatsiya i nomenklatura, 1981). Inset shows an additional sampling of "tholeiitic glasses" obtained in 1 atm experiments of more than 24 h duration, with olivine, plagioclase, and high-Ca pyroxene (+Spl) present on the liquidus. This assemblage coexists only with melts containing at least 8% MgO.

within the DBMS, compare them with those obtained earlier, print them, and, if necessary, run the calculation procedure again with new initial data or model parameters. This paper will not consider examples of creating melt-mineral geothermometers, analyzing redox equilibria, or calculating the water content in quench glasses: the reader is referred to papers specially devoted to these subjects (Ariskin and Nikolaev, 1996; Ariskin et al., 1996; Almeev and Ariskin, 1996; Nikolaev et al., 1996). Here we illustrate the efficiency of the INFOREX system using new options that make it possible to analyze the pattern of cotectic lines depending on pressure and temperature in certain petrologically important projections. Analysis of Ol-Opx-Cpx Cotectic Melts at High Pressure For example, consider the results of high-pressure experiments on Mg-rich basalts and peridotites. The query is formulated as follows: we need to find runs conducted on samples of natural and similar synthetic ultrabasites and basites at pressures of 10-20 kbar (dry conditions) in the temperature range 1250-1450œC. Additional search attributes include: experiment duration must be at least 12 h, and the experiment products must be high-magnesia glasses (Mg# > 0.65) whose compositions correspond to equilibrium with the lherzolite assemblage Ol-Opx-Cpx + Spl. In response to
PETROLOGY Vol. 5 No. 1 1997

this query, the INFOREX database produced 75 experiments matching these criteria: these runs represent 13 studies, conducted between 1977 and 1996 (Mysen and Kushiro, 1977; Takahashi and Kushiro, 1983; Fujii and Scarfe, 1985; Takahashi, 1986; Falloon and Green, 1987; Falloon et al., 1988; Bartels et al., 1991; Kinzler and Grove, 1992; Hirose and Kushiro, 1993; Falloon et al., 1994; Baker and Stopler, 1994; Baker et al., 1995; Kushiro, 1996). The Petrological Problem is to construct a generalized equation describing the compositional shift of the Ol-Opx-Cpx-Liq cotectic melt as a function of temperature and pressure. To accomplish this, we chose the Auxiliary Calculations option (Fig. 3) and recalculated the 75 glass compositions to the Ol, Pl, and Qtz components according to the procedure of Walker et al., (1979). This is the routine procedure used by petrologists to project multicomponent compositions onto pseudoternary diagrams (in this case, Ol-Pl-Qtz) so as to visualize them on a plane. The INFOREX program enables the user to process these data in a linear model of the form:1 T, œC = aP, kbar + b ln X Ol + c ln XPl + d, (1) where a, b, c, and d are regression coefficients. The regression parameters and their standard deviations cal1

For multiphase equilibria, such descriptions (sublinear models) are extremely common among petrologists, particularly when developing computer programs for calculating phase equilibria (Longhi, 1991, 1995; Kinzler and Grove, 1992; Yang et al., 1996).


32 Developing geothermometers: 1. 2. 3. 4. Olivine-melt Plagioclase-melt Pyroxene-melt Spinel-melt

ARISKIN et al. Calculation of water solubility in melt: 1. (Burnham, 1994) 2. (Dixon, Stolper, Hollway, 1995) 3. (Almeev and Ariskin, 1996)

PETROLOGICAL CALCULATIONS

Major-element recalculations and projection of glass compositions on ternary diagrams using the method: (Walker et al., 1979) 1. Ol-Pl-Qtz 2. Ol-Pl-Cpx 3. Ol-Cpx-Qtz (Grove, 1993) 1. 2. 3. 4. 5. 6. Ol-Cpx-Qtz Ol-Pl-Cpx Ol-Pl-Qtz Qtz-Ab-An Cpx-Ab-An Pl-Cpx-Qtz

Calculation of Fe3+/Fe2+ in melt: (Sack et al., 1980) (Kilinc et al., 1983) (Kress, Carmichael, 1988) (Borisov and Shapkin, 1989) (Mysen, 1989) (Kress, Carmichael, 1991) (Nikolaev et al., 1996) 7. Tholeiitic series 8. Subalkaline series 9. Alkaline series 1. 2. 3. 4. 5. 6.

Fig. 3. Block diagram showing the main functions of the Petrological Calculations module.

Tcalc, œC 1500
% Pav = +2.7 kbar 60 40 20 0 -12 -6 0 6 12 Pexp - Pcalc, kbar

1500 (a)

1400

1400 n = 75 Tav = +22œC

% Pav = +1.3 kbar 60 40 20 0 -12 - 6 0 6 12 Pexp - Pcalc, kbar

(b)

n = 60 Tav = +11œC 1300 10-20 kbar, including: (Mysen, Kushiro, 1977) (Falloon, Green, 1987) (Falloon et al., 1988) (Falloon et al., 1994) 1300 Texp, œC 1400 1500 1300 10-11 kbar 12-13 kbar 15-16 kbar 20 kbar 1200 1300 Texp, œC 1400 1500

1200

Fig. 4. Comparison of experimental and calculated temperatures of equilibria Ol-Opx-Cpx-melt at pressures of 10-20 kbar. Calculations were performed according to equation (1) using parameters in the table: (a) total data set of glass compositions extracted from the database; (b) the same data set excluding the experiments of Mysen and Kushiro (1977) and Falloon et al., (1987- 1994) (see the main text of the paper). The histograms represent the accuracy of solving the inverse problem for pressure.

culated for the compositions retrieved from the base are presented in the table. These parameters can be used to determine the accuracy with which equation (1) reproduces the values of experimental temperatures and to assess whether this

equation can be used for barometry at a known melt composition and equilibrium temperature. These data are shown in Fig. 4a and, at first sight, indicate that the accuracy is fairly high: the average deviations of the calculated temperatures and pressures from the experiPETROLOGY Vol. 5 No. 1 1997


INFOREX INFORMATION RETRIEVAL SYSTEM

33

Regression coefficients for the equation of high-pressure equilibrium Ol-Opx-Cpx-melt (1) and equation of amphibole liquidus under water-saturated conditions (2) Parameters a b c d Ol-Opx-Cpx-melt, P = 10-20 kbar, í = 1250-1450œC n = 75 8.14 7.91 - 82.60 1171.49 (1.39) (23.84) (15.56) (37.99) n = 60 8.65(0.77) 34.86(13.30) - 99.10(10.05) 1170.89(21.40) Amphibole-melt, P = 2-6.9 kbar, í = 800-1000œC n = 59 - 71.47(9.63) - 159.28(38.75) 668.93(50.44) 1317.39(65.32)

Note: Standard deviations (1) are given in parentheses.

mental ones are + 22.4œC and + 2.75 kbar, respectively. However, a close examination of 15 points with maximum deviations (45-87œC and 5-10 kbar) reveals that 13 of them represent results obtained in one laboratory (Falloon and Green, 1987; Falloon et al., 1988; Falloon et al., 1994), and the remaining two relate to early experiments on peridotite melting (Mysen and Kushiro, 1977), which were later revised on several occasions. In our opinion, these discrepancies may point to the presence of systematic, possibly methodical, differences in the results obtained in the Australian laboratory (Falloon et al., 1987-1994) from the data of other investigators (Takahashi and Kushiro, 1983; Fujii and Scarfe, 1985; Takahashi, 1986; Bartels et al., 1991; Kinzler and Grove, 1992; Hirose and Kushiro, 1993; Baker and Stolper, 1994; Baker et al., 1995; Kushiro, 1996). This conclusion does not indicate unambiguously which experiments yield the most accurate results, but practical considerations suggest that in constructing the liquidus equation (1), it would be useful to exclude the data of those 15 experiments that show the
Pl
100
1250

greatest deviations during the solution of the inverse problem. This opportunity is provided by a special option of the INFOREX system, enabling all experiments representing a particular study to be filtered out (Meshalkin and Ariskin, 1996). Resorting to this option, we excluded from consideration the data of Mysen and Kushiro (1977), Falloon and Green (1987), Falloon et al., (1988), and Falloon et al., (1994); the regression parameters for compositions extracted with this filter and obtained in the 60 "more accurate" experiments are also given in the table. These results show better agreement between the data and appreciably increased precision in solving the inverse problem (Fig. 4b). We have discussed this example in detail because it graphically demonstrates the potential of the INFOREX database as an expert system. Statistical processing of large amounts of information without carefully examining the data is likely to produce inadequately accurate dependencies. The INFOREX system

Experimental data: Mg# > 0.65 1250 T 1450 > 12 h 10-11 kbar 12-13 kbar 15-16 kbar 20 kbar Calculated isotherms: 10 kbar 20 kbar
0 25 50 75

1350

1350 1350 1375

1300 1400

1375 1375 1400 1425

10
20

kbar kbar

Ol 0

25

50

75

100 Qtz

Fig. 5. Composition of experimental glasses and calculated isotherms of high-pressure equilibrium Ol-Opx-Cpx-melt in the "barometric" projection Ol-Pl-Qtz. Recalculation to normative components was performed using the method of Walker et al., (1979). Attempts to graphically project natural compositions onto this diagram may lead to errors in estimating pressure of about 5-7 kbar. PETROLOGY Vol. 5 No. 1 1997


34 Qtz
100

ARISKIN et al.

8 85 00œC 9 0œ 95 00œC C 0œ 10 00 C œC

Experimental data: 800ϑ 850ϑ 900ϑ 950ϑ 1000ϑ
50

Calculated isotherms
75

excess water (vapor phase) subject to two additional conditions: (1) the data on the composition of the melt are available, and (2) the composition includes the final phase assemblage of high-calcium amphibole (hornblende). No constraints are imposed on the experiment duration. In the INFOREX database, these match criteria are met by 158 experiments published in 10 papers. The distribution of pressures is as follows: 66 experiments were carried out at pressures of 1-2.5 kbar, 63 at 3- 5.2 kbar, and 29 at 7-10 kbar. These experiments bracket the temperature range 675-1050œC, with glass compositions varying from basaltic andesite to rhyolite (52-80% SiO2). The melts retrieved correspond to equilibria with various assemblages, which include-- in addition to amphibole--plagioclase, olivine, orthopyroxene, augite, magnetite, ilmenite, biotite, apatite, and some other minerals. To further analyze these data, we selected the most representative studies (Beard and Lofgren, 1991; Luhr, 1990; Sisson and Grove, 1993) conducted at pressures of 2 kbar (28 experiments), 4 kbar (19 experiments), and 6.9 kbar (12 experiments). The petrological problem was to use this set to construct the liquidus equation for amphibole and estimate the effect of temperature and pressure on the composition of a melt in equilibrium with hornblende. For this purpose, we again chose the Auxiliary Calculations option (Fig. 3) and analyzed nine methods of recalculating glass compositions in order to find a projection, in which the desired relationships are seen most clearly. The best result was obtained with Grove's method, which was used to project the compositions onto the Qtz-An-Ab (quartz-anorthite-albite) plane; this is not surprising, if we consider that the experimental glasses are, on average, of dacite composition. After that, the data of 59 selected experiments were processed according to the linear model: T, œC = a ln P, bar + bXQtz + cXAn + d , (2)

25

0

50 25

75

100

2 kbar
0

Ab 0

25

50 mol %

75

100 An

Fig. 6. Composition of experimental glasses and calculated isotherms of amphibole-melt equilibrium under water-saturated conditions (P = 2 kbar). The glass compositions were projected onto the quartz- anorthite-albite plane using normative components calculated according to the method of Grove (1993).

allows the user to experiment with matching various data by constructing and analyzing regression equations (Ariskin and Nikolaev, 1996; Ariskin et al., 1996). The above analysis also has practical significance. In Fig. 5, the compositions of 60 high-pressure cotectic melts Ol-Opx-Cpx are projected onto the Ol-Pl-Qtz plane. Petrological interest in this projection arises from its barometric potentialities, because some investigators have repeatedly demonstrated that the olivine- pyroxene equilibrium lines are appreciably shifted toward the Ol corner with increasing pressure (Takahashi, 1986; Kinzler and Grove, 1992; Hirose and Kushiro, 1993; Baker et al., 1995); this shift is schematically shown in Fig. 5. Using equation (1), we calculated the equilibrium isotherms Ol-Opx-Cpx-Liq at pressures of 10 kbar (1250 and 1300œC) and 20 kbar (1350 and 1400œC) and superposed them on this scheme. The data presented show that, in attempts to graphically project the natural or calculated compositions of basaltic melts onto this diagram without considering the temperature of the assumed equilibrium with residual lherzolite, the error in estimating the pressure may be as high as 5-7 kbar. Calculation of Isotherms of Melt Saturation in Amphibole To provide another illustration of using the INFOREX database, we shall analyze the results of water-saturated experiments. The purpose of our query is to find all runs conducted on natural and synthetic samples at pressures of 1-10 kbar in the presence of

where a , b, c, and d are regression coefficients (table). During the solution of the inverse problem according to equation (2), the experimental values of temperature and pressure were reproduced within +19.4œC and +1111 bar (average deviations); this appears to be a good result, if we consider the large amount of methodical uncertainties associated with hydrous experiments and glass quenching. Figure 6 shows five isotherms calculated at a pressure of 2 kbar according to equation (2) using parameters in the table. These data clearly demonstrate that our liquidus equation for amphibole faithfully reproduces the shift of the melt compositions into the field enriched in quartz and albite with decreasing temperature. The estimates of water content in the melts according to the models of Burnham (1994) and Almeev and Ariskin (1996) vary within 4.2-6.0 and 5.5-7.2 wt % H2O, respectively.
PETROLOGY Vol. 5 No. 1 1997


INFOREX INFORMATION RETRIEVAL SYSTEM

35

CONCLUSION The INFOREX database with its built-in procedures of processing experimental information is a powerful new tool for petrological investigations. With this system, subliquidus experiments are searched and analyzed in a matter of seconds, and a user who is experienced at using the base will have to spend only a few hours to develop the most accurate mineral-melt geothermometers or liquidus equations for multiphase assemblages. The petrologist now has access to the results of about 7000 experiments on melting synthetic and natural samples, and about 10 000 compositions of experimental phases are available for testing geothermometers and various empirical dependencies. A new capability for working with experimental data is the calculation of water concentration in glasses for which there are no direct determinations of H2O content. The database is systematically updated annually. At the same time, it is somewhat disconcerting to see that the INFOREX system is being introduced into petrological research at an unreasonably slow pace. Scientists still tend to use traditional computer models of fractionation, melting, and other algorithms designed to solve certain specialized petrological problems. The question of the accuracy and applicability of the programs is commonly ignored or regarded as the developer's sole responsibility. Such an approach to computer modeling appears to be inadequate. As we see it, any direct application of computer models to natural rocks and minerals is possible only after these models are tested on experimental data covering that field of magma compositions that is of interest to each particular petrologist or geochemist. The INFOREX system offers unique opportunities to accomplish this task. ACKNOWLEDGMENTS This work was supported by the Russian Foundation for Basic Research, projects nos. 96-05-64231 and 96-07-89054. REFERENCES
Almeev, R.R. and Ariskin, A.A., Mineral-Melt Equilibria in a Hydrous Basaltic System: Computer Modeling, Geokhimiya, 1996, no. 7, pp. 624-636. Ariskin, A.A., and Nikolaev, G.S., Distribution of Fe3+ and Fe2+ between Chrome-Spinellid and Basaltic Melt as a Function of Composition, Temperature, and Oxygen Fugacity, Geokhimiya, 1995, no. 8, pp. 1131-1139. Ariskin, A.A. and Nikolaev, G.S., An Empirical Model for the Calculation of Spinel-Melt Equilibrium in Mafic Igneous Systems at Atmospheric Pressure: I. Chromian Spinels, Contrib. Mineral. Petrol., 1996, vol. 123, pp. 282-292. Ariskin, A.A., Barmina, G.S., Meshalkin, S.S., Nikolaev, G.S., and Almeev, R.R., INFOREX-3.0: A Database on Experimental Phase Equilibria in Igneous Rocks and Synthetic SysPETROLOGY Vol. 5 No. 1 1997

tems. II. Data Description and Petrological Applications, Comput. Geosci., 1996 (in press). Ariskin, A.A., Bouadze, K.V., Meshalkin, S.S., and Tsekhonya, T.I., INFOREX-3.0: A Database on Experimental Studies of Phase Relations in Silicate Systems, Am. Mineral., 1992, vol. 77, pp. 668-669. Ariskin, A.A., Frenkel, M.Ya., Barmina, G.S., and Nielsen, R.L., COMAGMAT: A FORTRAN Program to Model Magma Differentiation Processes, Comput. Geosci., 1993, vol. 19, pp. 1155-1170. Baker, M.B., Hirschmann, M.M., Ghiorso, M.S., and Stolper, E.M., Compositions of Near-Solidus Peridotite Melts from Experiments and Thermodynamic Calculations, Nature, 1995, vol. 375, pp. 308-311. Baker, M.B. and Stolper, E.M., Determining the Composition of High-Pressure Mantle Melts Using the Diamond Aggregates, Geochim. Cosmochim. Acta, 1994, vol. 58, pp. 2811-2827. Bartels, K.S., Kinzler, R.J., and Grove, T.L., High-Pressure Phase Relations of Primitive High-Alumina Basalts from Medicine Lake Volcano, Northern California, Contrib. Mineral. Petrol., 1991, vol. 108, pp. 253-270. Beard, J.S. and Lofgren, G.E., Dehydration Melting and Water-Saturated Melting of Basaltic and Andesitic Greenstones and Amphibolites at 1, 3, and 6-9 kb, J. Petrol., 1991, vol. 32, pp. 365-401. Borisov, A.A. and Shapkin, A.I., A New Empirical Equation Relating the Fe3+/Fe2+ Ratio in Natural Melts to Composition, Oxygen Fugacity, and Temperature, Geokhimiya, 1989, no. 6, pp. 892-898. Burnham, C.W., Development of the Burnham Model for Prediction of the H2O Solubility in Magmas, Rev. Mineral., 1994, vol. 30, pp. 123-129. Dixon, J.E., Stolper, E.M., and Holloway, J.R., An Experimental Study of Water and Carbon Dioxide Solubilities in Mid-Ocean Ridge Basaltic Liquids. Part 1: Calibration and Solubility Models, J. Petrol., 1995, vol. 36, pp. 1607-1631. Falloon, T.J. and Green, D.H., Anhydrous Partial Melting of MORB Pyrolite and Other Peridotite Compositions at 10 kbar: Implications for the Origin of Primitive MORB Glasses, Mineral. Petrol., 1987, vol. 37, pp. 181-219. Falloon, T.J., Green, D.H., Hatton, C.J., and Harris, K.L., Anhydrous Partial Melting of a Fertile and Depleted Peridotite from 2 to 30 kb and Application to Basalt Petrogenesis, J. Petrol., 1988, vol. 29, pp. 1257-1282. Falloon, T.J., Green D.H., and Jaques, A.L., Refractory Magmas in Back-Arc Basin Settings: Experimental Constraints on a Lau Basin Example, Mineral. Mag., 1994, vol. 58A, pp. 263-264. Fujii, T. and Scarfe, C.M., Composition of Liquids Coexisting with Spinel Lherzolite at 10 kbar and the Genesis of MORBs, Contrib. Mineral. Petrol., 1985, vol. 90, pp. 18-28. Ghiorso, M.S. and Sack, R.O., Chemical Mass Transfer in Magmatic Processes: IV. A Revised and Internally Consistent Thermodynamic Model for the Interpolation and Extrapolation of Liquid-Solid Equilibria in Magmatic Systems at Elevated Temperatures and Pressures, Contrib. Mineral. Petrol., 1995, vol. 119, pp. 197-212. Grove, T.L., Corrections to Expressions for Calculating Mineral Components in "Origin of Calc-Alkaline Series Lavas at Medicine Lake Volcano by Fractionation, Assimilation and Mixing" and "Experimental Petrology of Normal MORB near


36

ARISKIN et al. Meshalkin, S.S., Ariskin, A.A., Barmina, G.S., Nikolaev, G.S., and Almeev, R.R., Development of an Experimental Database on Crystal-Melt Equilibria in Igneous Rocks: INFOREX System (Version 3.0), Geokhimiya, 1996, no. 2, pp. 99-105. Mysen, B.O., Relations between Structure, Redox Equilibria of Iron, and Properties of Magmatic Liquids, Physical Chemistry of Magmas. Advances in Physical Geochemistry, Perchuk, L.L. and Kushiro, I., Eds., New York: Springer, 1991, vol. 9, pp. 41-98. Mysen, B.O. and Kushiro, I., Compositional Variations of Coexisting Phases with Degree of Melting of Peridotite in the Upper Mantle, Am. Mineral., 1977, vol. 62, pp. 843-856. Nielsen, R.L., Simulation of Igneous Differentiation Processes, Rev. Mineral., 1990, vol. 24, pp. 65-105. Nikolaev, G.S., Borisov, A.A., and Ariskin, A.A., Calculation of the Ferric-Ferrous Ratio in Magmatic Melts: Testing and Additional Calibration of Empirical Equations for Various Magmatic Series, Geokhimiya, 1996, no. 8, pp. 713-722. Sack, R.O., Carmichael, I.S.E., Rivers, M., and Ghiorso, M.S., Ferric-Ferrous Equilibria in Natural Silicate Liquids at 1 bar, Contrib. Mineral. Petrol., 1980, vol. 75, pp. 369-376. Sisson, T.W. and Grove, T.L., Experimental Investigations of the Role of H2O in Calc-Alkaline Differentiation and Subduction Zone Magmatism, Contrib. Mineral. Petrol., 1993, vol. 113, pp. 143-166. Takahashi, E., Melting of a Dry Peridotite KLB-1 up to 14 GPa: Implications on the Origin of Peridotitic Upper Mantle, J. Geophys. Res., [Solid Earth Planets], 1986, vol. 91, pp. 9367-9382. Takahashi, E. and Kushiro, I., Melting of a Dry Peridotite at High Pressures and Basalt Magma Genesis, Am. Mineral., 1983, vol. 68, pp. 859-879. Walker, D., Shibata, T., and DeLong, S.E., Abyssal Tholeiites from the Oceanographer Fracture Zone: II. Phase Equilibria and Mixing, Contrib. Mineral. Petrol., 1979, vol. 70, pp. 111-125. Weaver, J.S. and Langmuir, C.H., Calculation of Phase Equilibrium in Mineral-Melt Systems, Comput. Geosci., 1990, vol. 16, pp. 1-19. Yang, H.-J., Kinzler, R.J., and Grove, T.L., Experiments and Models of Anhydrous, Basaltic Olivine-Plagioclase-Augite Saturated Melts from 0.001 to 10 kbar, Contrib. Mineral. Petrol., 1996 (in press).

the Kane Fracture Zone: 22œ-25œN, Mid-Atlantic Ridge," Contrib. Mineral. Petrol., 1993, vol. 114, pp. 422-424. Hirose, K. and Kushiro, I., Partial Melting of Dry Peridotites at High Pressures: Determination of Compositions of Melts Segregated from Peridotite Using Aggregates of Diamond, Earth Planet. Sci. Lett., 1993, vol. 114, pp. 447-489. Kilinc, A., Carmichael, I.S.E., Rivers, M., and Sack, R.O., The Ferric-Ferrous Ratio of Natural Silicate Liquids Equilibrated in Air, Contrib. Mineral. Petrol., 1983, vol. 83, pp. 136-140. Kinzler, R.J. and Grove, T.L., Primary Magmas of MidOcean Ridge Basalts, 1: Experiments and Methods, J. Geophys. Res., [Solid Earth], 1992, vol. 97, no. 5, pp. 6885-6906. Kinzler, R.J. and Grove, T.L., Primary Magmas of MidOcean Ridge Basalts, 2: Applications, J. Geophys. Res., [Solid Earth], 1992, vol. 97, no. 5, pp. 6907-6926. Klassifikatsiya i nomenklatura magmaticheskikh gornykh porod: Spravochnoe posobie (Classification and Nomenclature of Igneous Rocks. Handbook), Bogatikov, O.A., Mikhailov, N.P., and Gon'shakova, V.I., Eds., Moscow: Nedra, 1981. Kress, V.C. and Carmichael, I.S.E., Stoichiometry of the Iron Oxidation Reaction in Silicate Melt, Am. Mineral., 1988, vol. 73, pp. 1267-1274. Kress, V.C. and Carmichael, I.S.E., The Compressibility of Silicate Liquids Containing Fe2O3 and the Effect of Composition, Temperature, Oxygen Fugacity, and Pressure on Their Redox States, Contrib. Mineral. Petrol., 1991, vol. 108, pp. 82-92. Kushiro, I., Partial Melting of a Fertile Mantle Peridotite at High Pressures: An Experimental Study Using Aggregates of Diamond, Geophysical Monograph 95, Basu, A. and Hart, S., Eds., Am. Geophys. Union, 1996, pp. 109-122. Longhi, J., Comparative Liquidus Equilibria of Hypersthenenormative Basalts at Low Pressure, Am. Mineral., 1991, vol. 76, pp. 785-800. Longhi, J., Liquidus Equilibria of Some Primary Lunar and Terrestrial Melts in the Garnet Stability Field, Geochim. Cosmochim. Acta, 1995, vol. 59, pp. 2375-2386. Luhr, J.F., Experimental Phase Relations of Water- and Sulfur-saturated Arc Magmas and the 1982 Eruptions of El Chicon Volcano, J. Petrol., 1990, no. 5, pp. 1071-1114. Meshalkin, S.S. and Ariskin, A.A., INFOREX-3.0: A Database on Experimental Phase Equilibria in Igneous Rocks and Synthetic Systems: I. Datafile and Management System Structure, Comput. Geosci., 1996 (in press).

PETROLOGY

Vol. 5

No. 1

1997